48 Do Deserving Subjects Not Receive Compensation? II 49 The Ethical Basis for Compensation 3 Any policy on the possible compensation of injured research subjects must take into account many practical factors: the cost of such a program, the program's vulnerability to abuse, technical difficulties in admin-istration, and the political influence of those who would benefit and those whose budgets might be adversely affec-ted. Behind these practical concerns, however, loom some basic ethical questions: Is there a moral responsibility to compensate research subjects for medical bills, lost wages, and other out-of-pocket costs that are a direct result of injuries sustained in research conducted or sponsored by the Federal government? Is the investiga-tor's obligation limited to informing subjects of the risks? Or is the subject's consent insufficient to guarantee that no wrong is done? Is compensation for injury required as a matter of justice? If not, might it be morally important because of beneficial effects on society, or because it provides the opportunity to realize an ethical ideal other than justice? Persons of good will differ in their beliefs concerning the moral importance of compensation of injured research subjects, and this division of opinion over the ethical issues explains some of the lack of consensus as to the need for a value of a compensation program. This chapter addresses the logic of the major distinctly ethical arguments voiced in favor of, and in opposition to, compensation for injured research sub-jects. The arguments are not definitive and thus will not overcome all doubts raised about research injury compensation. Rather, this exposition is intended to indicate the range of opinion reviewed by the Commission and to provide the reasoning behind the Commission's conclusions. Preceding page blank 51 Compensating for Research Injuries: Chapter 3 A Question of Justice The basic case for a program of compensation for injured subjects can be stated briefly: Medical and scientific experimentation, even if carefully and cautiously conducted, carries certain inherent dangers. Experimentation has its victims, people who would not have suffered injury and disability were it not for society's desire for the fruits of research. Society does not have the privilege of asking whether this price should be paid: it is being paid. In the absence of a program of compensation of subjects, those who are injured bear both the physical burdens and the associated financial costs. The question of justice is why it should be these persons, rather than others, who are to be expected to absorb the financial, as well as the unavoidable human costs of the societal research enter-prise which benefits everyone. The argument in favor of compensating injured subjects proceeds from a simple rule of thumb for determining the justice of the distribution of burdens and benefits in contexts like that of medical research: those who receive the benefits should be those who undertake the risks. In some medical research, the subjects are patients who volunteer precisely because the experiment offers the greatest chance of cure; prospec-tive benefits outweight prospective risks and distribu-tive justice is not a major concern. when, however, as is often the case in research with human subjects, those who bear the risks are not the direct beneficiaries of the research, it is felt that the scales of justice are out of balance. Institutional Review Boards and other protective mechanisms have been designed to ensure that the balance is thrown off-center as little as possible, consistent with the goal of permitting promising research to continue. Compensation may be regarded as a means of restoring the balance after the fact, when the residuum of risk not eliminated by the protective devices had eventuated in injury. Like the protective devices, compensation is a further means of limiting the burdens borne by individual subjects in research. The Argument from Fairness. The ethical norm underlying most of the literature favoring a program of compensation is that of fairness, a key element in the concept of justice. One formulation of a principle of fairness was provided to the Commission following the philosophers Hart and Rawls: If there is a "mutually beneficial scheme of Compensating for Research Injuries: Chapter 3 52 social cooperation," then "a person who has accepted the benefits of the scheme is bound by a duty of fair play to do his part and not to take advantage of the free benefits by not cooperating." 1/ Stated as such, this is a principle of distributive justice: it dictates an assignment of benefits and burdens that is held to be just. This principle is invoked, for example, by moral and political theorists to justify the extraction of taxes from those who may never have signed an actual agreement to contribute to the state's budget. The idea is simply that if one benefits while others take their turns, then one too must take his or her turn. This notion of fairness appears to an ideal of reciprocity in social relations, is discussed in Chapter One of this Report. In the research context, it is the government, representing the society, which must "take its turn." The research subject has contributed by exposing himself to risks and the government must do its part. Professor James Childress formulated for the HEW Task Force a related principle which called for compensation if: (1) The injured party accepts or is compelled to accept a position of risk...(2) The activity is for the benefit of society...(3) Society, through the govern- ment or its agencies, conducts, sponsors, or mandates the practice in question. 2/ If these features are present, then whether or not the injured party was a volunteer, [t]he moral principle of fairness creates a societal obligation to this participant, who can claim as his right not merely consideration of damages but compensation at least for major injuries. The obligation is voluntarily ________________ 1. Bernard Boxill, Consent and Compensation, (1980), Appendix C to this Report at 16 quoting John Rawls, Legal Obligation and the Duty of Fair Play, in Sidney Hook, ed., LAW AND PHILOSOPHY, New York University Press, New York (1964) at 9-10. 2. James Childress, Compensating Injured Research Subjects: I. The Moral Argument, 6 HASTINGS CTR. REP. 21.24 (December 1975). 53. Compensating for Research Injuries: Chapter 3 incurred by the society, through its establishment, endorsement, or mandate of the practice in question, and its acceptance of the practice in question, and its acceptance of the individual's participation, whether it drafts, encourages, or merely accepts him. This obligation is based on the relationship between the parties in question, not on the fact that society through biomedical research wrongfully injured the participant (which would have been a matter of reparative justice). It reflects the moral principle of fairness. 3/ Professor Childress's contribution seems to have been incorporated into the Task Force's official view, as expressed in the ethics chapter of its final Report: [B]ecause society is both the beneficiary and sponsor of research, compensatory justice ('that form of justice which seeks to redress injury even when no fault or blame is associated with the injury') may come into play for the redress of injuries suffered by persons in connection with biomedical or behavioral research conducted, supported, or regulated by the Federal Government. 4/ Another important precedent is the Veterans' Compensation for Service-Connected Disability or Death Program. The obligation here is toward those who have entered into a special relationship of service to the American society, and who are, therefore, entitled to special compensation in the event of a service-connected disability. In short, the American society has recognized a special obligation to compensate veterans for injuries sustained in connection with their service since society is both sponsor and beneficiary of their services. 5/ ________________ 3. Id. (emphasis in the original; footnote omitted). 4. HEW Secretary's Task Force on the Compensation of injured Research Subjects, REPORT, U.S. Department of Health, Education, and Welfare, Washington (1977)(hereinafter cited as TASK FORCE REPORT) at VI-4. 5. Id. at VI-2-3 Compensating for Research Injuries: Chapter 3 54 Further, soldiers,...could sustain compensable injury--even serious injury or death--without anyone being guilty of negligence or any other tort. [S]ociety, through its governmental agents, has intervened in the lives of the injured individuals, and therefore society may be said to have an obligation to repair (so far as possible) injury done to individuals, whether they are volunteers or draftees, in connection with their service to society. 6/ The Task Force cautioned that "the analogy...is by no means perfect," but reported being "impressed by the obligation on the part of society to provide compensation..." 7/ The Task Force's desire to base its recommendation on an appeal to fairness has evident appropriateness and appeal. Though these principles of fairness do not command universal assent among contemporary moral or political theorists, they have received considerable support in leading theories of distributive justice. They provide a moral basis for government mechanisms necessary for the harmonious functioning of large and complex societies, which must proceed in the absence of an actual contract and which would be crippled if denied the means for dealing with the problems of "free riders" who would take advantage of others' contributions to the public good. The point of the argument from fairness, then, is that the potential beneficiaries of medical research--which includes the entire citizenry--ought not have a "free ride" at the expense of injured research subjects. If the human costs of research are low--say a matter of a little time and inconvenience--than the allocation of costs is not a serious ethical - - -. When, however, higher costs are occasionally imposed, as in the case of injured subjects, the question "Who pays?" becomes important. Research subjects are already doing more than their share, namely by the fact of having volunteered. Those who are injured bare the greatest burden of all. According to this view, certainly, they are the most appropriate parties to have to bear the financial costs of injury. Let those costs be shouldered by the potential beneficiaries who have contributed neither (ILLEGIBLE) nor health to the research effort. The society outht to meet this expense, through (ILLEGIBLE) action if necessary, in order to compensate the injured subjects. _________________ 6. Id. at VI-3 7. Id 55. Compensating for Research Injuries: Chapter 3 . REMAINDER OF PAGE IS ILLEGIBLE Compensating for Research Injuries: Chapter 3 56 straightforward; the subject is told of the risks; he consents to joining the experiment even though those risks are present, and he has not been promised any compensation. Why, then would society be obligated to compensate in the case of injury? The Task Force, after much debate, took a strong stand on this question: Informed consent in the research setting functions as a recognition of and a protection for a person's integrity and autonomy, but does not imply a waiver of the right of the person to compensation in the event of injury.... Even if a subject perfectly understands a research procedure and agrees to participate in that procedure, the subject's consent does not, in and of itself, include explicitly or implicitly, a waiver of compensation. (V)olunteers may give inferred consent to participation in biomedical and behavioral research without thereby surrendering the right to be compensated should injury occur. 10/ But, if so, the Task Force stated, one ought to respect the autonomy of the potential subject, it is at least initially difficult to understand why there is a moral requirement to compensate a subject who consents to participation without the expectation of compensation in case of injury. If respect for autonomy requires the subject's consent, why would that same respect not also require that the subject be permitted to agree to shoul-der the risk of injury without the possibility of compen-sation? The Task Force, it must be recalled, was speaking of a consent form (and process) that - - - - listed the risks and benefits of procedures, together with a state-ment of certain rights of the patient having nothing to do with compensation. Perhaps, it could be said of that consent form, and of the process of obtaining consent which the form records, that the signing and consenting did not amount to an assumption of responsibiity for risk. Shortly after the Task Force's report, however, institutions receiving Federal funds for research were told to include an explicit statement on their policy of providing or not providing medical care and other __________________ 10. TASK FORCE REPORT, supro note 4, at - - (ILLEGIBLE) 57. Compensating for Research Injuries: Chapter 3 compensation. 11/ Thus the consent form became some-thing closer to an explicit assumption of risk. And the forms could be made even clearer on this score--for example, with the addition of a sentence such as: "In signing this form, I knowingly and freely assume all risks attendant to the - - - negligent conduct in this experiment and do not expect, and will not seek to hold others liable for compensation in case of injury not resulting from negligence." In this latter instance, if not at present, it is ________________________ 11. 43 FEDERAL REGISTER 31550 (November 2, 1978), now codified at 45 CFR 46.116(a)(o)(1981). FOR REFERENCE SEE (16bb06.gif) 58 Compensating for Research Injuries: Chapter 3 59 there is no injustice done to anyone by failing to compensate injured subjects. 12/ Fairness vs. Consent. These two perspectives on the question of whether failure to compensate injured subjects of medical research is unjust thus lead to quite different conclusions. The argument from fairness holds it to be unfair to impose the financial costs of injury upon those physically injured in the course of altruistic service to the community as research subjects. The second perspective stresses the freedom of citizens to volunteer their security as well as their convenience. As long as subjects make their choices freely, no injustice is done if only those subjects are accepted who assume responsibility for the cost of injuries. Which of these two perspectives ought to be adopted in the case of research-related injuries? The Commission recognizes that each is worthy of serious consideration, and further recognizes that the divergence of views on the matter of compensation of subjects is but one instance of a divergence in thinking about social justice generally. The fairness argument, stated in isolation, seems particularly convincing. It is, intuitively, quite unfitting that a person already injured should be saddled with attendant costs; and it is equally unfitting that potential subjects be asked to agree to assume those costs as a condition of being enrolled in reserch. This view ofjustice prevails in many other contexts. Soldiers and other Federal employees, for example, are not asked to waive compensation if injured on the job, even though it might be possible to recruit persons for these posit-ions even were such a waiver required. Indeed, even private employers are not permitted to ask employees to waive coverage under Workmen's Compensation programs. The question of what would constitute a just distribution of burdens and benefits is decided in these other con-texts before it is asked whether a waiver could be ob-tained; what matters is less whether an employee would agree to be denied compensation than whether this should even be asked. --------------------- 12. This is not to argue that compensation would violate any right of prospective patients to take chances. The regulations governing use of human subjects are in many respects protective of subjects regardless of their willingness to accept risks. The argument here is rather, that the government is not being unjust in being unprotective. 60. Compensating for Research Injuries: Chapter 3 When one fails to compensate injured subjects, though fiscal resources are available, one does not display the virtue of charity. Indeed, since the need arises because these subjects have displayed their concern for others by becomming subjects, a failure to compensate is not only uncharitable but unjust. A defin-itive answer would, it seems, require the Commission to choose between the rival views of justice, one emphasi-zing the achievement of an equitable pattern of distri-bution of the benefits and costs of research, the other stressing the transactions of free agents, whatever the resulting distribution. On the former view, the govern-ment has a strict obligation to compensate injured sub-jects, and its present failure to do so constitutes an injustice. On the latter view, uncompensated, injured subjects suffer no injustice so long as they freely and knowingly assumed responsibility for these costs before joining the experiment in which they were injured. Failure to compensate would then at most constitute a deficit in charity or benevolence. Several serious reservations about the consent that is obtained from subjects make the latter view less convin-cing to the Commission. First, the consent argument's appear to the notion of a gift freely offered would be strongest if the reserch subject were offered at the time he or she agreed to participate the alternatives of either having or not having compensation available should injury result. 13/ If subjects then reject the promise of compensation, they would appear to wish to donate 13. A contrary-to-fact hypothetical might help to clarify this point. Suppose that a private insurance company were willing to sell insurance policies to individuals for individual research projects and found a means to do which did not involve large transaction costs. The policies would compensate the subjects in case of injury, just as the program recommended by the Task Force would do. Imagine that a given subject is first asked to volunteer for a medium-risk study, and at the same time asked to buy his own insurance policy. The terms of the request, then, are that the subject donate his time and trouble, and in addition he pay out the (say) additional $1.50 that the insurance company charges for insuring him. The researcher accepts no volunteers who are unwilling to donate the $1.50 along with the time and trouble. One may suppose further that the researcher has no problem in finding enough people of good will and adequate pocket money for the experiment. The terms of the subjects agreement in this hypothetical would be perfectly and obviously just, in Compensating for Research Injuries: Chapter 3 61 both their time and their security. But in the absence of such an offer of future compensation, one cannot conclude that the consent of subjects who wish to aid research indicates their desire to make the additional gift of their security. Moreover, if injured subjects would accept compensation were it offered, then it would appear that they did not wish to make the gift of their security. This is not to say that consent could not be valid without an offer of compensation, but it does suggest that an element of capitalizing on subjects' desire to help science may sometimes occur. Many subjects will have an altruistic desire to aid research by offering their time, together with a personal desire not to put their security at risk. If the research investigator and the government can afford to offer compensation for injury, but refuse to do so knowing that they will still obtain volunteers, they thereby exploit the altruistic motivation of volunteers: when a subject gives the willing gift of his or her time, the unwilling gift of his or her security is extracted as well. If exploitation is a form of unfair taking advantage of another, then the government or researcher may act unfairly in asking for volunteers while refusing them the possibility of compensation for injury. view of those who construe consent as involving assump-tion of risk. Perhaps not everyone would be willing to volunteer for experiments if they had to buy an individual insurance policy as a condition for joining. But if they did, they will not have been treated unjustly. Again, if it is morally acceptable for them to volunteer their time and trouble, it must be morally acceptable for them to volunteer the extra $1.50. And if the were insured by a private insurer, there would be no apparent need for any government compensation program, even though the subjects may be injured in an experiment designed to benefit society generally and conducted or sponsored by the government. The contrariness-to-fact of this hypothetical deserves emphasis. The Commission has investigated the possibility of providing subjects with the chance to purchase individual insurance policies at the time of enrollment in research projects. The Commission's consultants have reported that such a scheme would be utterly impractical. Thus, if no compensation program is undertaken by research institutions or by the government, subjects will continue to be uninsured for lost wages, and for those having inadequate health insurance, for medical bills resulting from research injuries. 62. Compensating for Research Injuries: Chapter 3 In particular cases, there may be no ground for concern over the moral sufficiency of consent--for example, when an intelligent, financially secure, educated adult agrees to undertake a small accurately estimated risk of minor harm of known character. But rules concerning compensa-tion of subjects will not, as a practical matter, be able to distinguish these simple cases from the more difficult ones and the conditions of less-than-ideal consent are present often enough that the difficulties should be taken into account in formulating policy. In many experiments, the risks are not well known. A treatment may be so new that the pattern of adverse reaction or side effects has not been established, nor will it be possible to determine whether a given object is at especially high risk for these harms. Remote risks of serious harms are especially likely to attend experi-mental procedures. Consent in such cases is unnecessar-ily somewhat blind, and true appreciation of risk is doubtful, even for ideally competent subjects. Further, there is accumulating evidence that people do not perform well in calculating the expected utilities of events with small probabilities of occurrence. 14/ Events of different orders of magnitude of probability may be given the single rating of "unlikely". Assigning full respon-sibility to the subject for assumption of remote, but serious risks of research thus takes on the air of exploitation of known weaknesses. In the case of remote risks, when subjects agree to participate without compensation for injury, their gift of security is a small one because the substantial harm that might ensue is discounted for its low probability. For the unfortunate few subjects for whom the harm later materializes, however, an uncompensated injury dramatic-ally increases the size of their gift to the research enterprise. While they agreed at the outset to bear the injury without compensation should it occur, one may reasonably conclude that they did not expect it to occur and did not truly intend to make this larger gift. 15/ __________________ 14. Amos Iversky and Daniel Kahneman. The Framing of Decision and the Psychology of Choice, 211 SCIENCE 453 (1981). 15. Thus, injuries that arise because of unanticipated risks would seem to be more deserving of redress than those that arise from risks the subject knew about. One might wish, therefore, to experiment with a program in which eligibility depends on the type of risk manifested, but such a division would appear to be too complicated. Instead, it seems more sensible simply to regard all Compensating for Research Injuries: Chapter 3 63 Viewing the exchange in this light helps to account for some lingering sense that even allowing for the subject's valid consent to participate without compensation for injury, it is unfair for them to have to bear such a substantial burden. It may also help explain the strong obligation felt by many people (including many research-ers) to provide compensation in order to "repay" this gift (as noted in the discussion of the gift relationship in Chapter One), since the gift is not only large but probably more than the subjects initially intended to make. To these difficulties may be addded a host of stan-dard complaints about the consent process. Patients who are subjects are sometimes despite recitations of sub-jects' rights, in fear of displeasing their care-givers and hence in a dependent and unfree relationship. Pa-tients are often agitated and distraught because of the very medical condition that qualifies them for an exper-iment. Moreover, a significant number of subjects, including children, the severely retarded, and the senile, are simply incompetent. The propriety of altru-ism-by-proxy which may be asked of these subjects is not established, even for the small risks permitted by the HHS regulations. It can be presumed that under the current regulatory system, consent to treatment in most research involving human subjects meets the standards upon which society insists for valid transactions in other contexts, such as commerce. Indeed, the knowledge, competence, and inde-pendence of the contracting parties in society generally are seldom scrutinized as closely as are those qualities of potential subjects in biomedical research. Neverthe-less, it is fitting to use a high moral standard in the research context. The goals of human health and well-being that motivate research ought to be reflected as well in a higher moral standard than the caveat emptor of the marketplace. Additional Reasons for Compensation In addition to arguments based upon justice, there are at least two other moral grounds for a program of serious injuries to be, in some very real sense, "unanticipated" by those who suffer them. Like many other factors, the issue of unanticipated risks serves as a reminder of the ethical problems with placing too heavy reliance on the notion of "voluntary, informed consent" in the sense of a forced waiver of any claim for recompense for research injuries. 64. Compensating for Research Injuries: Chapter 3 compensation. Appropriate Regard for Patient-Subjects' Well Being. Justice is not the only standard of morality. There are acts, or failures to act, that are surely wrong even though they violate no one's rights and are not instances of injustice. Acts can be mean-spirited, selfish, cheap, irresponsible, though they comply with the rules of conduct required by justice. Ordinary English does not provide a precise vocabulary in which to distinguish these sorts of wrongs, nor does moral theory itself speak with one voice--in part because the classifications are matters of substance as well as semantics. Still, it is worthwhile to make a rough distinction between considerations of justice and other, still important, moral considerations. As regards the present suibject, the most important such considera-tion would be the distress of anyone injured in the course of research who was faced with large medical bills and loss of income due to research-related disability. Such instances of need suggest the simplest of all arguments for a program of compensation: the serious need of those few injured in research can be met without untoward expense or difficulty by a program administered through research institutions. If, as Henry Beecher wrote, medical experimentation must be "ethical in its inception," 16/ then, one may add, it must also be ethical in its consequences. A program of compensation for research-related injuries would help to ensure that the consequences of research would be good ones rather than bad. In this view, not to adopt such a program would be a selfish and cheap or even irresponsible disregard of patient-subjects' well being. Of course, this argument is unduly simple. The claim advanced is not in itself sufficient to establish the rightness of enacting a compensation program, although it may lend support to a decision that rests on a stronger moral claim. But it depends on showing that a serious need of injured subjects can be met without imposing disproportionate burdens on others. One is, after all, not considered irresponsible for failing to do something heroic, only for failing to meet a great need without great cost to oneself. Public Conceptions of Justice. Although the moral sensibilities of the public on this issue are not known, there are certain indirect indications that lack of 16. Henry K. Beecher, Ethics and Clinical Research, 274 NEW ENG. J. MED. 1354,1360 (1966). Compensating for Research Injuries: Chapter 3 65 compensation of injured subjects is, or would be, seen as wrong. A television documentary on injured subjects, for example, stressed the lack of compensation as one of the wrongs perpetrated. 17/ And the government gave consid-erable publicity to its efforts to locate and compensate the victims in certain widely reported cases of research-related injury. 18/ Finally, many individual researchers provide the immediate medical care needed by injured sub-jects, presumably because they feel some sort of moral obligation, even though they are (as a matter of their "contract" with the subjects) under no legal obligation to provide such free treatment. Thus, whether or not the moral argument in favor of an obligation to compensate is compelling, many members of the public subscribe to it. Even those who are not themselves convinced that failing to compensate injured subjects is a true injustice may not wish to see their government involved in an action that has even the appearance of an injustice. If, then, research injuries come to public attention yet remain uncompensated, public support for research with human subjects might be reduced. Failure to redress injuries, particularly if they are large, could be taken as evidence that those who direct the research enterprise do not have as strong an interest in the well-being of their research subjects as they do for their scientific endeavors. The resulting erosion of confidence that scientists and the public share an identity of goals could result in a climate of opinion inimical to research and, indeed, to all risk-taking for public benefit. This result might satisfy those whose sole agenda is the pro-tection of subjects from risks, but it would work to the detriment of society as a whole. Research on human subjects, if done as part of a broad attack on disease, is of necessity risk-laden. The public, if it is to support this enterprise, must find the human costs tolerable. It will be more likely to do so if the costs are widely shared through a program of compensation of subjects who are injured in research. The moral importance of medical research--a value worthy of respect and protection--thus adds weight to a conclu-sion in favor of compensation for research injuries. ________________ 17. Mission: Mind Control, produced by Paul Altmeyer, American Broadcasting Company, January 30, 1979. 18. See, e.g., James H. Jones, BAD BLOOD, Free Press, New York (1961) at 217-19. 66. Compensating for Research Injuries: Chapter 3 Obligations to Subjects in Therapeutic Research A crucial issue is whether compensation is owed to all subjects whose medical condition worsens in some respect in the course of research. The ethical arguments already developed are premised upon some notion of sacri-fice, of giving a benefit to the whole. This situation is met most clearly by subjects in "nontherapeutic re-search" who undergo procedures unconnected with their own condition (indeed, they are often 'normal volunteers'). At first sight, "therapeutic research"--that is, an experiment designed to gain generalizable knowledge about a medical intervention by employing it under controlled conditions with a group of patient-subjects--would appear to be disqualified ethically because participating sub-jects may be seeking benefit for themselves as patients rather than being solely vessels through which knowledge can flow to others. This view would treat therapeutic research as an example of ordinary (albeit innovative) therapy for purposes of compensation; patients who become subjects are seen as taking no greater risk than they would otherwise face in dealing with their illness. The distinction between therapeutic research and innovative therapy is a narrow one. In innovative ther-apy, a physician is simply trying something new to bene-fit the patient often because existing remedies have failed. In therapeutic research, a physician-investigator follows a research protocol in order to produce generalizable knowledge through testing out a medical intervention that it is hoped, may be of benefit. This goal creates at least a potential conflict of interest on the part of the physician-investigator, over and above any hazards inherent in the particular study design. Each patient who is also a research subject is exposed to the possibility that the procedures most conducive to his recovery will be altered so that the research program can best be carried out (for example, that random assignment will remove the pssibility of access to certain procedures). 19/ The patient could avoid this jeopardy ____________________________ 19. The considerations set forth here do not require compensation of the subjects who experience poorer outcomes in randomized trials in cases in which all arms had equal chances of therapeutic success. This will be true even if most or all of the enrollees in one of the arms do much less well than those in the other arms, and even if those who do less well would have done better had they received standard therapy outside of the experiment. The reason is that the subjects, at the time of entry Compensating for Research Injuries: Chapter 3 67 by seeking the innovative therapy outside of the research context, where this is available. Where the patient in-stead decides to become the subject of therapeutic re-search, he or she makes a gift (however slight) of some of his or her security to the larger society. Moreover, some therapeutic research will not offer a patient his or her best chance for recovery. It is, furthermore, quite common for therapeutic experiments to involve procedures, especially tests, that are performed only for scientific studies may be intended to be "thera-peutic" for the patient-subject, but compensation for injuries would rest on the same footing as for nonthera-peutic research: the patient is making a contribution to society. Conclusions A program of compensating for research injuries will help to equalize the burdens of progress in this field that would otherwise fall very unevenly on people. Competing conceptions of justice lead to contrary conclusions about whether justice requires that such a program be adopted. Fairness argues for compensation, but the alternative argument from the consent of free agents denies this. Although, for both philosophical and practical reasons, it is not persuaded by the latter argument, the Commis-sion does not hold that compensation is a basic right (such as the right to a fair trial), to be provided regardless of cost, practicality, or other policy considerations. The arguments for compensation of injured subjects on fairness grounds simply do not establish so strict an obligation. Several additional considerations add weight to the arguments for compensation beyond the notion of fairness, both in ethical theory and in the public perception of the government's stance. First, the moral claims of charity and generosity ought not to be ignored--although the obligations they create have inherent limits, for one into the study, did not take positions of added risk: as far as could be ascertained, each arm was as likely as were the others to turn out to be the best one. (Where the arms of the randomized trial are not of equal expected therapeutic value of course, the arguments in favor of compensating might apply.) Hence there would be no sacrifice imposed on the subjects by the study, nor would there be an act of altruism involved in enrolling, unless there were added dangers from any incidental procedures, added to the patients' treatment solely for research reasons. 68. Compensating for Research Injuries: Chapter 3 not judged wrong in failing to be heroic. Second, the frailties of the consent process, especially with regard to small risks or serious harm, weigh heavily as a reason for added "protection" in the form of after-the-fact recompense at least for any serious harms experienced by subjects. 20/ The Commission concludes that compensation of injured subjects is appropriate to the research enterprise. A program to assure compenstion is thus a desirable policy goal for a just and compassionate government, both as the sponsor of most biomedical and behavioral research and as the means through which soci-ety acts on matters of common interest, such as the search for new biomedical discoveries. Whether a program should be adopted depends on the need for it (Chapter Four) and the alternative means (Chapter Five), and upon whether the program's transaction costs, vulnerability to abuse, or difficulty of administration would be dispro- 20. A further benefit of a general program of compensa-tion would be a partial amelioration of certain harms visited by unethical research practices. The overwhelm-ing majority of research subjects are recruited under honorable conditions. Research scientists, while having a research agenda in addition to any therapeutic rela-tionship with a potential subject, have generally been as solicitous of patient welfare as they would be for their own. The current review process is in place, however, because there have been exceptions. Only a few cases have come to light, but there does not exist a means of monitoring the actual consent process closely enough to deter a scientist determined to break the rules. Fur-ther, studies of the incidence of injury have relied heavily on reporting by those who would have caused the injuries. Thus there is some cause for concern over injury due to participation in research under fraudulent conditions, as when potential benefits are exaggerated or risks are undisclosed. There are existing legal remedies for harms visited upon subjects through deception and fraud, and it would be convenient for the Commission to be able to formulate its recommendations without having to consider these cases. It is reasonable, however, to ask whether the existing legal protection is adequate, and, if not, to suggest policies which might help to protect subject. A general program of compensation for subjects injured in research would provide a partial remedy for the problem of injuries to subjects recruited through deception, and the Commission counts this as a point in favor of a compensation program. Compensating for Research Injuries: Chapter 3 69 portionate to the ethical problems at which the program would be aimed (Chapter Six). Moreover, a full evalua-tion of the ethical arguments for and against compensa-tion for research injuries must take these practical con-siderations into account. Such practicalities are in themselves ethical matters because they turn on questions of the fair allocation of resources, the potential impact of a program on the rights of other parties, and horizon-tal equity with persons suffering from other misfortunes.